Server is a significant infrastructure for network architecture. In the server, a basis input/output system (BIOS) is an extremely important module, commonly. In booting, it requires to initialize various hardware devices according to a BIOS setting, whereby an operation system can operate the respective hardware devices after starting running.
Currently, a server system has mostly one BIOS chip disposed therein. If a power on self-test (POST) fails, it requires to repair or offline-update a firmware of the BIOS chip. This is inconvenient for users. If the server system has an extra BIOS chip additionally disposed therein, the server system can automatically switch to the other BIOS chip to normally activate the server system as long as either one of the BIOS chips fails to perform the POST or its firmware has a defect in itself. This provides a great convenience for a data center or an architecture employing a great amount of servers. In conventional skills, a desktop computer can also adopt the afore-described BIOS chip backup design but it requires manual intervention to achieve the BIOS chip switching effect.
In conventional designs, a baseboard management controller (BMC) disposed in the server system is applied to electrically connect a to-be-monitored BIOS chip with a serial peripheral interface (SPI) of the baseboard management controller, thereby carrying out monitoring and determining, by the baseboard management controller, whether or not the BIOS chip perform the POST, normally. However, there have following problems if two BIOS chips are electrically connected to the baseboard management controller at the same time. That is, how to identify the two BIOS chips for determining which one of the BIOS chips can be used to active the server system normally.
Therefore, there is a need to provide a novel server system for solving above problems.